502

REFERENCE: Reilly CA, Crouch DJ, Yost GS. Quantitativeanalysis of capsaicinoids in fresh peppers, oleoresin capsicum andpepper spray products. J Forensic Sci 2001;46(3):502–509.

ABSTRACT: Liquid chromatography-mass spectrometry wasused to identify and quantify the predominant capsaicinoid ana-logues in extracts of fresh peppers, in oleoresin capsicum, andpepper sprays. The concentration of capsaicinoids in fresh pepperswas variable. Variability was dependent upon the relative pungencyof the pepper type and geographical origin of the pepper. Non-ivamide was conclusively identified in the extracts of fresh peppers,despite numerous reports that nonivamide was not a natural product.In the oleoresin capsicum samples, the pungency was proportionalto the total concentration of capsaicinoids and was related by a fac-tor of approximately 15,000 Scoville Heat Units (SHU)/�g of totalcapsaicinoids. The principle analogues detected in oleoresin cap-sicum were capsaicin and dihydrocapsaicin and appeared to be theanalogues primarily responsible for the pungency of the sample.The analysis of selected samples of commercially available pepperspray products also demonstrated variability in the capsaicinoidconcentrations. Variability was observed among products obtainedfrom different manufacturers as well as from different product lotsfrom the same manufacturer. These data indicate that commercialpepper products are not standardized for capsaicinoid content eventhough they are classified by SHU. Variability in the capsaicinoidconcentrations in oleoresin capsicum-based self-defense weaponscould alter potency and ultimately jeopardize the safety and healthof users and assailants.

KEYWORDS: forensic science, pepper spray, capsaicin, LC/MS

For centuries, people have used concentrated extracts of peppers(oleoresin capsicum) or the dried fruits to prepare spicy foods(1–4). Other historical and current uses of pepper products includeneurobiological research (5), weight loss (6–8), local anesthesia(5,9), anti-microbial defense (10), anti-inflammation preparations(5,11), and recently for the production of self-defense and less-than-lethal (LTL) weaponry (3,12). The extensive use of peppersand pepper extracts for such diverse purposes emanates from thepresence of capsaicinoids in the pepper.

The term “capsaicinoids” describes a group of pungent chemicalanalogues found in hot peppers (Capsicum annum and C.frutescens) (1–4). There are five naturally occurring capsaicinoids:capsaicin, dihydrocapsaicin, nordihydrocapsaicin, homocapsaicin,and homodihydrocapsaicin (1–4). The most abundant and potentanalogues in peppers (and consequently pepper extracts) are cap-saicin and dihydrocapsaicin (1–4,13). Nordihydrocapsaicin, homo-

capsaicin, and homodihydrocapsaicin are also present, but gener-ally contribute little to the total capsaicinoid concentration andpungency of the pepper (1–4,13). Nonivamide, or “synthetic” cap-saicin, exhibits the same pungency as capsaicin, but has not beenconclusively identified as a natural product (3,13). In fact, somescientists have concluded that detection of nonivamide in oleoresincapsicum was indicative of adulteration of the sample (3). The to-tal concentration of capsaicinoids in a pepper ranges from 0.1 to2.0% (dry weight) and depends upon the variety of the pepper, thegrowing conditions, and the time of harvest (1,3,4). In addition todifferences in the total capsaicinoid content, variations in the rela-tive proportions of the capsaicinoid analogues also occurs in response to the above criteria (1–4).

Capsaicinoids are the pharmacologically active and pain-pro-ducing components of the hot pepper (4). The characteristic chem-ical structure of capsaicin (or its analogues) contains a vanillamidemoiety (4-hydroxy-3-methoxybenzylamide) and an acyl chain con-taining 10 to 11 carbon atoms (Fig. 1) (1–4,12). Capsaicinoids pro-duce pain by stimulating the vanilloid receptor. This receptor is amolecular integrator of potentially noxious stimuli (e.g., low pHand high temperature) (5,11,14,15). Structure activity studies ofcapsaicin and other related compounds have demonstrated a strictrequirement for the vanillyl ring and an acyl chain of 8 to 12 car-bon atoms to manifest pungency (16–18). The natural capsaici-noids exhibit variable pungency due to differences in their abilityto promote membrane depolarization through binding to the vanil-lamide receptor.

The ability of the capsaicinoids to produce pain has promptedthe development of pepper sprays for the purpose of self-defense.Typically, pepper sprays weapons contain a 10% solution of oleo-resin capsicum diluted in a suitable solvent (e.g., methylene chlo-ride, trichloroethylene, isopropanol, freon(s), propylene glycol,ethanol, methanol, or dimethyl ether) and a gaseous propellant(usually N2 or CO2). Exposure to pepper sprays elicits an intensephysiological response that includes nociception, temporary blind-ness, lacrimation, disorientation, shortness of breath, and choking(19). The result of the exposure is temporary incapacitation of thevictim with minimal long-term side effects and/or toxicity (19,20).

Unfortunately, the manufacturers of oleoresin capsicum andself-defense weaponry employ few, if any, analytical measures todetermine the concentration of active ingredients in the product andto ensure consistent chemical composition. Variability in the con-centration of active ingredients might explain why pepper sprayshave been shown to be only 70% effective in discouraging attacksby aggressive individuals (20). The principal test for product com-position is a taste test to determine the product’s ability to elicit

Christopher A. Reilly,1 Ph.D.; Dennis J. Crouch,1 B.S., M.B.A.; and Garold S. Yost,1 Ph.D.

Quantitative Analysis of Capsaicinoids in FreshPeppers, Oleoresin Capsicum and Pepper SprayProducts

1 Center for Human Toxicology, Department of Pharmacology and Toxicol-ogy, 20 S. 2030 E., Rm 490, University of Utah, Salt Lake City, UT.

Received 31 May 2000; accepted 5 July 2000.

REILLY ET AL. • ANALYSIS OF CAPSAICINOIDS 503

pain. This test is known as the Scoville Organoleptic Test (21). Un-fortunately, the Scoville Organoleptic Test gives only a subjectivemeasure of potency (SHU) rather than a quantitative assessment ofthe capsaicinoid concentration (21). Since the concentration of ac-tive ingredient in pepper products (including pepper sprays) is onlysubjectively assessed, the potency, efficacy, and potential producttoxicity is difficult to predict (22).

In this study, we identified and quantified the individual capsai-cinoid analogues in various fresh peppers, oleoresin capsicums,and pepper spray weapons. Nonivamide was identified in five dif-ferent types of fresh peppers originating from different locations,as well as in all commercial pepper products. We also calculatedthe relationship between pungency and capsaicinoid concentrationin oleoresin capsicum as well as in a series of pepper-based self-de-fense products.

Materials and Methods

Chemical

Extreme caution must be used when handling capsaicinoidsand/or pepper products to prevent serious discomfort and/or in-jury. Capsaicin (60% purity), capsaicin (E-8-methyl-N-vanillyl-6-nonenamide) (98% purity), dihydrocapsaicin (8-methyl-N-vanil-lylnonanamide) (98% purity), and nonivamide (N-vanillyl-nonanamide) were purchased from Sigma Chemical Co. (St. Louis,MO). Nordihydrocapsaicin, E-homocapsaicin, and homodihydro-capsaicin were purified from oleoresin capsicum using HPLC.Vanillamine hydrochloride and octanoyl chloride were purchasedfrom Aldrich Chemical Company, Inc. (Milwaukee, WI). Oc-tanoyl-vanillamide was synthesized by condensation of vanil-lamine and octanoyl chloride as previously described (12). HPLC-grade methanol and reagent-grade n-butyl chloride were purchasedfrom Burdick and Jackson (Muskeegon, MI). Formic acid, sodiumphosphate, and sodium chloride were purchased from Mallinckrodt(Paris, KY). The oleoresin capsicum samples were obtained fromvarious spice vendors and the pepper sprays were purchased fromindependent distributors.

Analytical Methods

Identification and quantification of the individual capsaicinoidswas performed by liquid chromatography-mass spectrometry usinga Hewlett-Packard series 1100 LC/MSD (Agilent Technologies,Palo Alto, CA). The capsaicinoids were separated using a MetaSilBasic reversed-phase HPLC column (100 � 3.0 � 3�) (MetaChemTechnologies, Torrance, CA) and a stepwise gradient of methanoland distilled water containing 0.1% (v/v) formic acid. The massspectrometer was set to detect the M�H positive ions of octanoyl-vanillamide (m/z � 280), nonivamide (m/z � 294), nordihydro-capsaicin (m/z � 294), capsaicin (m/z � 306), dihydrocapsaicin(m/z � 308), homocapsaicin (m/z � 320), and homodihydrocap-saicin (m/z � 322).

Preparation of the analytical standards was achieved by weigh-ing the appropriate quantity of each analogue on a Cahn 4700 ana-lytical balance (Cahn Instruments, Cerritos, CA) and dissolving thecompounds in methanol. Standards were stored at �20°C and werestable for the duration of the study. Calibration curves, containingall analogues and 500 ng octanoyl-vanillamide (as an internal stan-dard), were constructed from 0 to 5000 ng. Calibration curves weregenerated by calculating the peak area ratio (obtained using HPChemstation software) of the analyte to internal standard. Linearregression analyses were performed using the least squaresmethod. Separate curves were used to quantify the low- (1 to 50ng), mid- (50 to 500 ng), and high- (500 to 5000 ng) range concen-trations of the analogues. Split curves were needed to more accu-rately quantify the concentration of the compounds in the samples.

Analysis of Fresh Peppers

Extracts of fresh peppers were prepared by homogenizing ap-proximately 5 to 50 g of fresh pepper (purchased from food storesin UT, NM, and MD) in 5 volumes of phosphate-buffered saline(50 mM phosphate buffer, pH 7.2 containing 1 M NaCl) using aTeflon/glass homogenizer. To decrease the sample size of the freshpeppers, the stem and a portion of the pericarp were removed; thestem and pericarp have been shown to contain only negligible con-

FIG. 1—Chemical structures of the capsaicinoid analogues and octanoyl-vanillamide (internal standard).

504 JOURNAL OF FORENSIC SCIENCES

centrations of the capsaicinoids (1,23,24). The homogenate was ex-tracted for 15 min at room temperature with n-butyl chloride. Theextract was centrifuged at 1000 � g for 10 min and the upper or-ganic layer was collected. The extraction process was repeated andthe organic layers were combined and evaporated to dryness undera stream of air at 40°C. The dried residues were reconstituted in200 �L n-butyl chloride:methanol (1:1). The reconstituted residueswere diluted 20-fold in n-butyl chloride:methanol (1:1) and a 5 �Laliquot was combined with 500 ng octanoyl-vanillamide. The sam-ples were evaporated to dryness, reconstituted with 100 �L of 70%methanol: 30% distilled H2O, and analyzed using LC/MS.

Analysis of Oleoresin Capsicum

Commercially available oleoresin capsicum samples were di-luted 200-fold in n-butyl chloride:methanol (1:1). A 5 �L aliquotof the diluted product was pipetted into a 13 � 100 mm silanizedglass tube and octanoyl-vanillamide (500 ng) was added as the in-ternal standard. The sample was evaporated to dryness, reconsti-tuted with 100 �L of 70% methanol: 30% distilled H2O, and ana-lyzed using LC/MS.

Analysis of Pepper Spray

Individual pepper spray canisters were vigorously shaken andcooled to �20°C overnight in a freezer. The sprays were then gen-tly discharged into a 16 � 100 silanized glass tube that had beenpreviously equilibrated to �80°C using dry ice. Cooling the tubes

with dry ice was necessary to prevent evaporation of the solventduring collection. The sample was immediately capped and thawedon ice. The sample volume was determined and the volatile com-ponents were permitted to evaporate at room temperature with gen-tle agitation. The sample volume was reestablished by adding n-butyl chloride:methanol (1:1). Samples were then diluted 40-foldin n-butyl chloride:methanol (1:1) and a 5 �L aliquot combinedwith octanoyl-vanillamide (500 ng). The samples were evaporatedto dryness, reconstituted with 100 �L of 70% methanol: 30%dH2O, and analyzed using LC/MS.

Results

Analysis of Capsaicinoids by LC/MS

Identification of the individual capsaicinoid analogues (see Fig.1) was achieved by LC/MS. The calibration curve for the assay waslinear from 1.0 to 5000 ng, but was split into three curves to moreaccurately calculate the concentration of the analytes in the sam-ples. Identification of the individual capsaicinoids was achieved bymass-selective detection as well as by retention time. Octanoyl-vanillamide (m/z � 280) eluted at approximately 8.4 min followedby nordihydrocapsaicin (m/z � 294; 11.8 min), nonivamide (m/z �294; ~12.6 min), capsaicin (m/z � 306; ~12.6 min), dihydrocap-saicin (m/z � 308; ~15.8 min), homocapsaicin (m/z � 320; ~16.1min), and homodihydrocapsaicin (m/z � 322; ~17.5 min). A typi-cal mass-chromatogram of a standard containing 500 ng of allanalytes and internal standard is shown in Fig. 2. The compound

FIG. 2—Representative reconstructed ion mass-chromatogram of a 500 ng standard mixture containing all capsaicinoids and internal standard (500ng). Octanoyl-vanillamide is represented by the panel labeled m/z � 280, nordihydrocapsaicin and nonivamide by m/z � 294, capsaicin by m/z � 306,dihydrocapsaicin by m/z � 308, homocapsaicin by m/z � 320, and homodihydrocapsaicin by m/z � 322. Specific retention times for the individual ana-lytes are presented in the results section. Data were collected as described under materials and methods.

REILLY ET AL. • ANALYSIS OF CAPSAICINOIDS 505

(m/z � 308) eluting at approximately 16.2 min has been tentativelyidentified as decanoyl-vanillamide, based on similarities in reten-tion time and mass when compared to a synthetic standard of de-canoyl-vanillamide (data not shown).

Analysis of Capsaicinoids in Fresh Peppers

The analysis of “unadulterated” pepper extracts demonstratedthe presence of all previously documented naturally occurring cap-saicinoids as well as nonivamide. The identification of nonivamidein peppers was confirmed by mass (m/z � 294) and retention time(approximately 12.6 min). Confirmation of the presence of non-ivamide in the peppers was achieved using collision-induced frag-mentation of the molecular ion by LC/MS/MS and obtaining theproduct ion spectrum (data not shown). Nonivamide was present inhabeñero, anaheim, red-chili, green-chili, and green bell pepper ex-tracts at a relative concentration range of 0.42 to 7.2% of the totalcapsaicinoid concentration, depending upon the variety of pepper.Nonivamide was not detected in extracts of yellow bell pepper andwas present in the green bell pepper at less than 0.025% of the con-centration in green-chili pepper. The concentrations of the capsai-

cinoid analogues in the six varieties of peppers are summarized inTable 1.

Analysis of Oleoresin Capsicum

As described, quantification of the individual capsaicinoid ana-logues in commercial preparations of oleoresin capsicum wasachieved by comparing the peak area ratio of the samples to the cal-ibration curves. The capsaicinoid analogue concentrations in sevendifferent oleoresin capsicum samples are summarized in Table 2.The absolute and relative concentrations of the individual capsaici-noids varied between samples. Figures 3A and 3B illustrate the cor-relation between SHU rating and the total (Fig. 3A) and individualcapsaicinoid analogue concentrations (Fig. 3B). The relationshipbetween total capsaicinoid concentration and SHU rating was ap-proximately 15 000 SHU/�g of total capsaicinoids. Data presentedin Fig. 3B indicate that capsaicin and dihydrocapsaicin were pri-marily responsible for the SHU rating. Their relationships were approximately 35 700 and 33 900 SHU/�g, respectively. The totaland relative analogue concentrations were consistent between oleoresin capsicum samples having the same SHU rating.

TABLE 1—Capsaicinoid analogue concentrations in extracts of fresh peppers.

NDHC† Nonivamide Capsaicin DHC† HC† HDHC†Variety of Pepper Total (�g/g)* (%) (%) (%) (%) (%) (%)

Yellow Bell 0.0018 � 0.0002 N.D. N.D. 37 � 4.0 63 � 5.0 N.D. N.D.(Utah)

Green Bell 0.0049 � 0.0004 13.1 � 0.9 7.2 � 0.4 27 � 2.0 52 � 7.0 N.D. N.D.(Utah)

Green-chili 19 � 2.0 22.4 � 0.2 2.7 � 0.2 21 � 6.0 46 � 2.0 0.9 � 0.4 7 � 2.0(Utah)

Anaheim 73 � 7.0 11.8 � 0.9 1.0 � 0.1 47 � 2.0 36 � 3.0 1.3 � 0.4 3.0 � 0.7(New Mexico)

Anaheim 87 � 5.0 6.3 � 0.8 0.42 � 0.07 60 � 9.0 30 � 4.0 2.6 � 0.5 1.4 � 0.3(Maryland)

Red-Chili 83 � 9.0 11 � 1.0 0.6 � 0.1 56 � 4.0 29 � 2.0 1.0 � 0.5 1.8 � 0.4(New Mexico)

Red-chili 59 � 6.0 9.6 � 0.4 1.2 � 0.1 50 � 3.0 27 � 1.0 6.1 � 0.4 6.2 � 0.5(Utah)

Habeñero 510 � 27.0 3.6 � 0.3 1.4 � 0.1 61 � 6.0 32 � 3.0 0.9 � 0.1 1.1 � 0.1(Utah)

* Approximately 5 to 50 g of fresh pepper was extracted as described under materials and methods. Data are representative of the mean � standard de-viation of triplicate analysis of pepper extracts.

† Abbreviations for Nordihydrocapsaicin (NDHC), Dihydrocapsaicin (DHC), Homocapsaicin (HC), and Homodihydrocapsaicin (HDHC).

TABLE 2—Capsaicinoid analogue concentrations for various oleoresin capsicum samples.

NDHC† Nonivamide Capsaicin DHC† HC† HDHC†Sample Identity‡ Total (�g/�L)* (%) (%) (%) (%) (%) (%)

OC1 0.8 � 0.2 16.7 � 0.3 1.8 � 0.4 38 � 1.0 36 � 2.0 7.0 � 3.0 N.D.OC2 14 � 1.0 20 � 2.0 1.9 � 0.2 36 � 4.0 38 � 3.0 2.0 � 0.1 1.9 � 0.2OC3 18.4 � 0.4 18.7 � 0.3 1.2 � 0.9 35.9 � 0.5 39.3 � 0.4 2.0 � 0.1 2.9 � 0.4OC4 63 � 1.0 11 � 1.0 3.4 � 0.5 33 � 2.0 48 � 3.0 1.4 � 0.2 2.9 � 0.7OC5 71 � 2.0 8.6 � 0.7 5.2 � 0.8 39 � 3.0 42 � 1.0 1.5 � 0.1 2.4 � 0.3OC6 67 � 3.0 6.3 � 0.4 5.5 � 0.2 47.8 � 0.9 35.8 � 0.7 1.7 � 0.8 2.1 � 0.1OC7 131 � 5.0 7.7 � 0.1 4.2 � 0.1 41.2 � 0.5 43.5 � 0.9 1.3 � 0.2 2.2 � 0.4

* Data are representative of the mean � standard deviation of triplicate analysis of the oleoresin capsicum sample.† Abbreviations for Nordihydrocapsaicin (NDHC), Dihydrocapsaicin (DHC), Homocapsaicin (HC), and Homodihydrocapsaicin (HDHC).‡ Represents oleoresin capsicum samples with the following SHU ratings: OC1 (2.0 � 104 SHU), OC2 (2.0 � 105 SHU), OC3 (5.0 � 105 SHU),

OC4-6 (1.0 � 106 SHU), OC7 (2.0 � 106 SHU). Although OC4-6 have the same SHU rating, they were obtained from different manufacturers of oleo-resin capsicum.

506 JOURNAL OF FORENSIC SCIENCES

Analysis of Capsaicinoids in Pepper Spray

Total and individual capsaicinoid analogue concentrations forseveral commercially available pepper spray products are summa-rized in Table 3. The total capsaicinoid concentrations were consis-tent with previous reports of the composition of pepper sprays (25).The capsaicinoid concentrations in pepper spray products from dif-

ferent product lots from the same manufacturer are also presented inTable 3. A comparison between the labeled and calculated SHU val-ues for the pepper sprays is shown in Table 4. Collectively, these datademonstrate that there were substantial differences in the total andrelative capsaicinoid concentrations in the pepper spray products.Also, the labeled SHU values of the pepper sprays were not consis-tent with the assayed values.

FIG. 3—(A) The relationship between Scoville Heat Unit rating and total capsaicinoid concentration in oleoresin capsicum. (B) The relationship be-tween Scoville Heat Unit rating and the concentration of individual capsaicinoid analogues in oleoresin capsicum; (�) nordihydrocapsaicin (y � 4.7 �10�6 x � 1.0), (�) nonivamide (y � 3.17 � 10�6 x � 0.3), (�) homodihydrocapsaicin (y � 1.5 � 10�6 x � 0.02), (�) capsaicin (y � 2.8 � 10�5 x �1.9), (�) dihydrocapsaicin (y � 2.95 � 10�5 x � 2.0), and (�) homocapsaicin (y � 8.5 � 10�7 x � 0.1). Data represent the mean � standard deviationof triplicate samples. Analysis of the capsaicinoids in oleoresin capsicum was performed as described under materials and methods.

REILLY ET AL. • ANALYSIS OF CAPSAICINOIDS 507

Discussion

The data demonstrate that capsaicin, dihydrocapsaicin, nordihy-drocapsaicin, homocapsaicin, homodihydrocapsaicin, and non-ivamide can be found in fresh peppers, oleoresin capsicum, andpepper spray products. However, the total and relative concentra-tions of the capsaicinoid analogues in the peppers and the pepperproducts were quite variable (Tables 1 to 3). In all samples, cap-saicin and dihydrocapsaicin constituted 60 to 90% of the total cap-saicinoid concentration, followed by nordihydrocapsaicin (2 to20%), homocapsaicin (1 to 5%), homodihydrocapsaicin (1 to 5%),and nonivamide (1 to 5%).

Capsaicin, nonivamide, and dihydrocapsaicin are the most pun-gent capsaicinoid analogues having 100, 100, and 75% relativepungencies, respectively (12,13,18). Nordihydrocapsaicin, homo-capsaicin, and homodihydrocapsaicin exhibit relative pungenciesthat range between 20 to 50% that of capsaicin (12,13,18). Becausenordihydrocapsaicin, homocapsaicin, and homodihydrocapsaicincontribute little to the relative pungency of the peppers or pepperproducts, their contribution to pungency is not discussed below.However, this does not imply that these analogues may not be important in the toxicological and/or pharmacological properties ofthe products.

To determine the source of variability in pepper sprays and ole-oresin capsicum, we first analyzed the capsaicinoid concentrationsin fresh peppers. In the extracts of fresh peppers, the total concen-tration of the capsaicinoids, as well as that of each analogue, wasvariable and dependent upon the variety and geographical origin ofthe pepper. The total capsaicinoid concentration in extracts of thefresh peppers was reflective of the relative “hotness” of the pepper(habeñero�anaheim � red-chili�green-chili�green bell�yellowbell). Capsaicin and dihydrocapsaicin were the most abundant cap-saicinoid analogues present in the extracts of fresh peppers andcomprised 60 to 90% of the total capsaicinoid concentration. Thesum of the concentrations of capsaicin and dihydrocapsaicin in theextracts of fresh pepper paralleled the relative pungency of the pep-per type and appeared to dictate the pungency of the fruit (see Table1). For example, extracts of habeñero peppers (the “hottest” pep-per) contained approximately 90% capsaicin and dihydrocapsaicinwhile extracts of the milder green-chili pepper contained about65% capsaicin and dihydrocapsaicin.

The immediate precursor to pepper sprays is oleoresin capsicum,the extracted product of hot peppers. As expected, we also ob-served variability in the total and relative concentrations of the cap-saicinoids in commercial preparations of oleoresin capsicum. Wefound that there was a strong relationship between the concentra-tions of the capsaicinoids in commercial preparations of oleoresincapsicum and the SHU rating of the product (Fig. 3A). Comparingthe individual capsaicinoid analogue concentrations to SHU ratingdemonstrated that the concentrations of capsaicin and dihydrocap-saicin were good predictors of the pungency of the oleoresin cap-sicum sample (Fig. 3B). For example, oleoresins having an SHUrating 5.0 � 105 SHU contained approximately 65% capsaicinand dihydrocapsaicin and samples with an SHU rating 1.0 � 106

consisted of �80% capsaicin and dihydrocapsaicin.Since both fresh peppers and oleoresin capsicum exhibited var-

iability in the concentrations of the capsaicinoids, we were not sur-prised find significant differences in the total and relative concen-trations of capsaicinoids of the pepper sprays. An interesting dis-covery was that variability was present in all commerciallyavailable pepper sprays even in samples from the same manufac-turer (Table 3). For example, two identical pepper spray productsfrom different lots and the same manufacturer exhibited strikingdifferences in total capsaicinoid concentration, 2.5 � 0.2 �g/�Lversus 13.5 � 0.1 �g/�L. Variability in product composition wasalso observed in products from different manufacturers that wereboth labeled as 10% oleoresin capsicum with a rating of 2.0 � 106

TABLE 3—Capsaicinoid analogue concentrations of various pepper sprays.

NDHC† Nonivamide Capsaicin DHC† HC† HDHC†Sample Identity Total (�g/�L)* (%) (%) (%) (%) (%) (%)

PS1‡ 16.0� 0.4 6.4 � 0.2 1.6 � 0.1 51 � 1.0 37.1 � 0.8 2.2 � 0.1 1.5 � 0.1PS2‡ 11.8 � 0.4 8.7 � 0.3 2.0 � 0.1 45 � 2.0 40 � 1.0 2.0 � 0.1 2.1 � 0.1PS3 0.95 � 0.03 5.7 � 0.2 1.8 � 0.1 52.4 � 0.9 37 � 1.0 2.4 � 0.3 0.8 � 0.1PS4 8.3 � 0.4 0.2 � 0.01 1.9 � 0.1 50 � 3.0 35 � 1.0 1.6 � 0.1 2.0 � 1.0PS5§ 2.5 � 0.2 9.7 � 0.6 1.9 � 0.1 41 � 3.0 41 � 3.0 2.2 � 0.2 2.5 � 0.9PS6§ 13.5 � 0.1 10.8 � 0.1 1.8 � 0.1 36.5 � 0.2 45.8 � 0.3 1.8 � 0.1 2.0 � 0.8PS7 13 � 1.0 9.6 � 0.9 2.2 � 0.2 42 � 5.0 42 � 4.0 1.8 � 0.2 1.9 � 0.8PS8�� 27 � 3.0 N.D. 100 � 10.0 N.D. N.D. N.D. N.D.PS9�� 32 � 1.0 N.D. 100 � 4.0 N.D. N.D. N.D. N.D.

* Total capsaicinoid concentrations (�g/�L) of dispensed pepper spray. This value does not represent the total capsaicinoid content in the canister. Dataare representative of the mean � standard deviation of triplicate analysis of the pepper spray product.

† Abbreviations for Nordihydrocapsaicin (NDHC), Dihydrocapsaicin (DHC), Homocapsaicin (HC), and Homodihydrocapsaicin (HDHC).‡, §, �� Represents pepper sprays from the same manufacturer, but different product lots.

TABLE 4—Calculated versus labeled SHU value for various pepperspray products.

Sample Total Labeled CalculatedIdentity (�g/�L)* SHU SHU

PS1 16.0 � 0.4 2.0 � 106 2.4 � 105

PS2 11.8 � 0.4 2.0 � 106 1.8 � 105

PS3 0.95 � 0.03 2.0 � 106 1.5 � 104

PS4 8.3 � 0.4 2.0 � 106 1.3 � 105

PS5 2.5 � 0.2 1.5 � 106 3.7 � 104

PS6 13.5 � 0.1 1.5 � 106 2.0 � 105

PS7 13 � 1.0 1.5 � 106 1.9 � 105

PS8† 27 � 3.0 1.0 � 106 9.0 � 105

PS9† 32 � 1.0 1.0 � 106 1.0 � 106

* Data are representative of the mean � standard deviation of triplicateanalysis of the pepper spray products.

† Calculated SHU values were determined using the formula SHU �[capsaicin] (�g/�L) � 33 000 (SHU/�g) since only nonivamide was pre-sent (nonivamide has similar pungency to capsaicin (13) and the concen-tration was not in range for the trend line for nonivamide).

508 JOURNAL OF FORENSIC SCIENCES

SHU. The two products had vastly different capsaicinoid concen-trations (0.95 � 0.03 �g/�L versus 16.0 � 0.4 �g/�L). Estimat-ing the SHU value based on the total capsaicinoid concentration ofthe pepper sprays suggested that the labeled SHU values weresometimes overstated by a factor of �100 times (see Table 4).

Trends suggesting that the relative analogue concentrations ofcapsaicin and dihydrocapsaicin were responsible for the SHU rat-ing of the product were apparent for both fresh peppers and oleo-resin capsicum samples. In pepper sprays, the relative concentra-tions of capsaicin and dihydrocapsaicin were �80%. This mayimply that the majority of the pepper sprays may have been pre-pared as dilutions of oleoresin capsicum having an SHU rating1.0 � 106. Unfortunately, dilution of oleoresin capsicums havingsimilar SHU rating does not appear to guarantee consistency in theamount of active ingredient in the pepper spray. Therefore, thevariability in the capsaicinoid concentrations of the pepper sprayscan be attributed to the inherent variability in the capsaicinoid con-centrations of different batches of oleoresin capsicum and ulti-mately the particular properties of the extracted peppers.

From our data, it is reasonable to conclude that manufacturers ofoleoresin capsicum-based pepper sprays do not standardize theirproduct for the concentration of capsaicinoids. This was not truefor samples PS8 and PS9 (Table 4). The manufacturer of this prod-uct used a known quantity of nonivamide as the active ingredientrather than a standard volume of oleoresin capsicum. Previouswork has demonstrated that nonivamide exhibits identical pun-gency to capsaicin based upon the Scoville Organoleptic Test(12,13,18). It also has a similar structure activity relationship usingdesensitization of the VR receptor as a model system to predictpungency (12–18). Therefore, products fortified with or composedsolely of nonivamide are likely to have similar efficacy to oleoresincapsicum-based pepper sprays for use as self-defense weapons.However, the studies reported here did not evaluate or compare thebiological efficacy of products.

Nonivamide is not commonly used to manufacture peppersprays because it is considered a “synthetic” and/or pharmaceuticalagent. Oleoresin capsicum is a natural product. However, peppersprays manufactured from oleoresin capsicum are not regulated.Thus far, the identification of nonivamide in fresh peppers has notbeen definitively confirmed and its existence as a natural productdebated (3). Some authors suggested that nonivamide did not occurnaturally because of the structure of its acyl chain (3). This argu-ment was supported by inconclusive analytical data (3). GC/MSmethods were ineffective at differentiating nordihydrocapsaicinand nonivamide since they have the same molecular weight(3,26–31). Standard HPLC methods did not chromatographicallyseparate capsaicin from nonivamide (3, 31–35). The method de-scribed here allowed us to uniquely identify and quantify non-ivamide in extracts of fresh pepper, oleoresin capsicum, andpeppers sprays. Nonivamide was present at a concentration thatrepresented up to 7.2% of the total capsaicinoid concentration(Tables 1–3).

Based on our work, several conclusions about the concentrationsof capsaicinoids in peppers as well as products manufactured fromextracts of peppers can be made. Because oleoresin capsicum is ob-tained by extracting fresh peppers, variability in the total and rela-tive capsaicinoid concentrations exists. As such, products madefrom the oleoresin capsicum (e.g., food products, pepper sprays,etc.) also exhibit variability. Differences in the concentrations ofactive components in pepper sprays may affect the quality, effi-cacy, and safety of these products (20,37,38). Discrepancies in theamount of active ingredient in the products may result in unpre-

dictable results when the products are used for self-defense or tosubdue suspects. Differences in the concentration of active ingre-dients of the products could also affect the potential toxicity of theproducts and could jeopardize the safety and health of individualswho are exposed to and/or depend upon the products for protection(20,37–39). Quantitative analytical methods are needed to deter-mine the concentration of capsaicinoids in pepper sprays. Imple-menting this objective quality control measure, as well as regulat-ing the formulation of pepper sprays, would substantially increasethe predictability of product potency, efficacy, and its potential tocause toxicity.

Acknowledgments

The authors would like to express their gratitude to Kate andCody Dwire at ChemArmor Inc., for helpful discussions and forproviding samples of various pepper sprays and oleoresin cap-sicum. We would also like to thank the various spice venders forproviding the samples of oleoresin capsicum. This work was su-pported, in part, by a grant from the National Institute of Standardsand Technology (Department of Commerce Contract#: 60NAN-BOD0006).

References

1. Govindarajan VS. Capsicum production, technology, chemistry, andquality. Part I: history, botany, cultivation, and primary processing. CRCCrit Rev Food Sci Nutr 1985;22:109–76.

2. Surh YJ, Lee SS. Capsaicin, a double edged-sword: toxicity, meta-bolism, and chemopreventive potential. Life Sci 1995;56:1845–55.

3. Cordell GA, Araujo OE. Capsaicin: identification, nomenclature, andpharmacotherapy. Annals of Pharmacol 1993;27:330–6.

4. Govindarajan VS, Sathyanarayana MN. Capsicum-production, technol-ogy, chemistry, and quality. Part V: impact on physiology, pharmacol-ogy, nutrition, and metabolism; structure, pungency, pain, and desensiti-zation sequences. Food Sci and Nutr 1991;29:435–74.

5. Szallasi A, Blumberg PM. Vanilloid (capsaicin) receptors and mecha-nisms. Pharmacol Rev 1999;51:159–212.

6. Henry CJ, Emery B. Effect of spiced food on metabolic rate. Hum NutrClin Nutr 1986;40:165–268.

7. Cameron-Smith D, Colquhoun EQ, Ye JM, Hettiarachchi M, Clark M G.Capsaicin and dihydrocapsaicin stimulate oxygen consumption in theperfused rat hindlimb. Int J Obes 1990;14:259–70.

8. Lim K, Yoshioka M, Kikuzato S, Kiyonaga A, Tanaka H, Shindo M, etal. Dietary red pepper ingestion increases carbohydrate oxidation at restand during exercise in runners. Med Sci Sports Exerc 1997;29:355–61.

9. McMahon SB, Lewin G, Bloom SR. The consequences of long-term top-ical capsaicin application in the rat. Pain 1991;44:301–10.

10. Jones NL, Shabib S, Sherman PM. Capsaicin as an inhibitor of thegrowth of the gastric pathogen Helicobacter pylori. FEMS MicrobiolLett 1997;146:223–7.

11. Holzer P. Capsaicin: cellular targets, mechanisms of action, and selec-tivity for thin sensory neurons. Pharmacol Rev 1991;43:143–201.

12. Nelson EK. Vanillyl-acyl amides. J Amer Chem Soc 1919;41:2121–30.13. Jones ECS, Pyman FL. Relation between chemical constitution and pun-

gency in acid amides. J Amer Chem Soc 1925;127:2588–98.14. Tominaga M, Caterina MJ, Malmberg AB, Rosen TA, Gilbert H, Skin-

ner K, et al. The cloned capsaicin receptor integrates multiple pain-pro-ducing stimuli. Neuron 1998;21:531–43.

15. Caterina MJ, Leffler A, Malmberg AJ, Martin AJ, Trafton J, Peterson-Zeitz KR, Klotzenberg M, Basbaum AI, Julius D. Impaired nociceptionand pain sensation in mice lacking the capsaicin receptor. Science2000;288:306–13.

16. Walpole CS, Wrigglesworth R, Bevan S, Campbell EA, Dray A, JamesIF, et al. Analogues of capsaicin with agonist activity as novel analgesicagents; structure-activity studies. 1. The aromatic “A-region.” J MedChem 1993;36:2362–72.

17. Walpole CS, Wrigglesworth R, Bevan S, Campbell EA, Dray A, JamesIF, et al. Analogues of capsaicin with agonist activity as novel analgesicagents; structure-activity studies. 2. The amide bond “B-region.” J MedChem 1993;36:2373–80.

REILLY ET AL. • ANALYSIS OF CAPSAICINOIDS 509

18. Walpole CS, Wrigglesworth R, Bevan S, Campbell EA, Dray A, JamesIF, et al. Analogues of capsaicin with agonist activity as novel analgesicagents; structure-activity studies. 3. The hydrophobic side-chain “C-re-gion.” J Med Chem 1993;36:2381–9.

19. Hyder K. Is CS the Wrong Solution? New Sci 1996;149:12–13.20. Hepburn JR, Griffin ML, Petrocelli M. Nonlethal weapons and use of

force. In: Office of Law Enforcement Standards (OLES). Status Report,1997.

21. Scoville WL. Note on Capsicum. J Am Pharm Assoc 1912;1:453.22. DiBartolomis MJ, Howd RA, Bankowaska J, Fan AM, Jackson RJ, Book

SA. Health risk evaluation of tear gas products containing oleoresin cap-sicum. Toxicologist 1993;13:417.

23. Huffman VL, Schadle FR, Villalon B, and Burns EE. Volatile compo-nents and pungency in fresh and processed jalapeño peppers. J Food Sci1978;43:1809.

24. Balbaa SL, Karawya MS, Girgis AN. The capsaicin content of capsicumfruits at different stages of maturity. Lloydia 1968;31:272.

25. Haas JS, Whipple RE, Grant PM, Andersen BD, Volpe AM, Pelkey GE.Chemical and elemental comparisons of two formulations of oleoresincapsicum. Sci Jus 1997;37:15.

26. Kosuge S, Furata M. Studies on the pungent principles of Capsicum:chemical constitution of the pungent principles. Agric Biol Chem1970;34:248–56.

27. Müller-Stock A, Joshi RK, Buchi J. Components of capsaicinoids:quantitative gas chromatographic determination of individual homo-logues and analogues of capsaicin in mixtures from a natural sourceand vanilylpelorgonamide as an adulteration. J Chrom 1971;63:281–7.

28. Jurenitsch J, Leinmüller R. Quantification of nonylic acid vanillamideand other capsaicinoids in the pungent principle of capsicum fruits andpreparation by gas-liquid chromatography on glass capillary columns. JChrom 1980;189:389–97.

29. Jurenitsch J, David M, Heresch F, and Kubelka W. Detection and iden-tification of new pungent compounds in capsicum fruits. Planta Med1979;36:61–7.

30. Lee K-R, Suzuki T, Kobashi M, Hasegawa K, and Iwai K. Quantitative

microanalysis of capsaicin, dihydrocapsaicin, and nordihydrocapsaicinusing mass fragmentography. J Chrom 1976;123:119–28.

31. Iwai K, Suzuki T, and Fujiwake H. Simultaneous microdetermination ofcapsaicin and its four analogues by using high-performance liquid chro-matography and gas chromatography-mass spectrometry. J Chrom1979;172:303–11.

32. Saria A, Lembeck F, Skofitsch G. Determination of capsaicin in tissuesand separation of capsaicin analogues by high-performance liquid chro-matography. J Chrom 1981;208:41–6.

33. Fung T, Jeffery W, Beveridge AD. The identification of capsaicinoids intear gas. J Forensic Sci 1982;27:812–21.

34. Weaver KM, Awde DB. Rapid high-performance liquid chromato-graphic method for determination of very low capsaicin levels. J Chrom1986;367:438–42.

35. Müller-Stock A, Joshi RK, Buchi J. Thin-layer and column chromato-graphic separation of capsaicinoids: study of the constituents of cap-sicum. J Chrom 1973;13:577–9.

36. Glinsukon T, Stitmunnaithum V, Toskulkao C, Buranawuti T,Tangkrisanavinont V. Acute toxicity of capsaicin in several animalspecies. Toxicon 1980;18:215–20.

37. Kaminski RS, Edwards SE, Johnson JW. In: Office of Law EnforcementStandards (OLES). Status Report, 1997.

38. Busker RW, Van Helden HP. Toxicologic evaluation of pepper spray asa possible weapon for the Dutch police force: risk assessment and effi-cacy. Am J Forensic Med Pathol 1998;19:309–16.

39. Granfield J, Onnen J, and Petty CS. Pepper spray and in-custody deaths.In: Science and Technology, International Association of Chiefs of Po-lice, Alexandria, VA. 1994.

Additional information and reprint requests:Dennis J. CrouchCenter for Human ToxicologyDepartment of Pharmacology and Toxicology20 S. 2030 E., Rm 490University of UtahSalt Lake City, UT 84112